Upload
caise2013vlc
View
312
Download
0
Tags:
Embed Size (px)
Citation preview
© Giannoulis
Jelena Zdravkovic, Janis Stirna, Martin Henkel Department of Computer & Systems Sciences (DSV), Sorckholm University, Sweden
Jānis Grabis Information Technology Institute, Riga, Technical University, Latvia
Modeling Business Capabilities and Context Dependent Delivery by Cloud Services
© Giannoulis
Outline
Problem
Capability and Other Key Concepts
Capability Meta-Model, Delivery in Cloud
Example Case
Capability Driven Development
Conclusion and Future Work
© Giannoulis
Problem
© Giannoulis
The Challenge • Changing business contexts
• Limited support for changes in non-
functional requirements
• Inability to model execution
contexts
• High cost of developing applications
for different contexts
• Gap between EM and MDD
• Limited cloud platform usage
© Giannoulis
Capability and Other Key Concepts
© Giannoulis
Key Concepts: Capability • In a business context, capability refers to the resources and
expertise that an enterprise needs to offer its functions. • Lately the notion of business capability has gained a growing
attention, due to a number of factors: it directs business investment focus, it can be used as a baseline for business planning, and it leads directly to service specification and design • Capability maps to IT deployments through IT architectures.
• The notions has been over time captured by Enterprise
Architecture, but neither in details modeled, nor formally linked to software services.
© Giannoulis
Key Concepts: Context • The notion of context refers to situational cognition; as such, it
is used to describe the conditions of a situation. • “any information that can be used to characterize the situation
of an entity” Where, who, what... Location, identity, activity, time… Modalities of agent’s state in AOP…
• Existing context categorizations set the focus to an entity, or
more specifically, to a user. In contrast, in our research, there is a need to model the context surrounding the delivery of business.
© Giannoulis
Key Concepts: Cloud • Cloud computing facilitates coherence of the resources and
economies of scale through its pay-per-use business model. • From the customer’s perspective the cloud technology offers a
means to increase capacity or add capabilities on the fly, without investing in new infrastructure, training new personnel, or licensing new software.
• So far, scalability challenges have been mainly addressed by providing targeted technical solutions.
• Our approach starts by setting up business requirements for the cloud using capabilities and context.
© Giannoulis
Key Relations
Enterprise and capability (Goal, KPI, Process)
Capability in context (Context, Context KPI)
Capability patterns (Process Variants, Resources)
Delivery in cloud (Cloud Models, Implementation)
© Giannoulis
Capability Meta-Model
Capability Meta-Model, Delivery in Cloud
Context Framework and Measurable Properties
Context Category
Relevance Availability Feature Time Location
Subjects Organization Customers Partners Competitors
What is subject doing?
Is subject available?
Characteristic or quantity of subject
When does subject perform process?
Where is subject located?
Objects Infrastructure Artefact Service
How is object used?
Is object available?
Characteristic or quantity of object
When is object used?
Where is object located?
Environment Regulations Standards Weather
What is the influence of environment?
Is environment concept available?
Characteristic or quantity of environment
When is environment concept applicable?
Where is environment concept located?
Example Case
Example Case - Goals
Example Case – Capabilities • Strategic planning providing the building operator and other
stakeholders with decision support for investment planning based on historical data and scenarios for prices and weather.
• Operational planning in passive mode providing operational planning based on template based energy audit and human assisted or manual input of data into DSS and manual input of temperature set points for the next 24 hours into the system used for controlling.
• Operational planning in active mode providing operational planning based on full energy audit, and daily update of sensor data and automatic transfer of set-points into the BEMS using the BACnet/IP protocol.
Example Case – Context Context Set for Capability: Operational planning in passive mode
Context Type Relevance Availability Feature Time Location
Subjects Weather data provider
Provides weather data
Yes, as cloud service
Date of last update
Every day Not relevant
Pricing data provider
Provides pricing data
Yes, as cloud service
Date of last update
Every day Not relevant
Building operator
Update energy usage and weather data
Yes Not relevant Every 24 hours At the building or remotely
Objects BEMS For data input
and output No Not relevant Not relevant At the building site
Sensors For building environment and weather
No Sensor readings Every 15 min At the building site
Example Case – Delivery in Cloud? Payment process. Variants: pay-per-use, one-time fee. • For cloud delivery it is rational to let the users of the DSS to
pay a monthly or yearly fee, since the use of a cloud platform will incur cost for the provider.
• For a local installation it can be assumed the organization buying the service will provide own hardware, thus allowing a one-time fee.
• In the EnRiMa case it is likely that other services (e.g. energy audits) will be performed by the organization providing the DSS, thus this points toward using a flexible pay-per-use fee.
Example Case – Delivery in Cloud? Deployment and Update process. Variants: local installation scripts, upload in cloud. • The deployment of software to the cloud and locally can differ,
thus to support both cloud delivery and local installations there is a need to have separate process variants for each.
• Some cloud platform, such as Amazon EC2 IaaS service, support the upload of pre-configured virtual machines, while others, such as Google App Engine PaaS service requires the service components to have specific format compliant with the platform.
• In the EnRiMa DSS case the DSS user interface services component is developed in a format that is easily transferrable to Google App Engine PaaS.
Capability Driven Development
Capability Driven Development Environment
Cloud Service
Capability design tool
Enterprise architecture
Context platform
SaaS
IaaS
PaaS C
apab
ility
del
iver
y ap
plic
atio
n
KPI
Adjustment algorithms
Modeling module
Rep
osito
ry o
f pa
ttern
s
Composition module
Integration module
Conclusion and Future Work
Conclusion We have proposed to support the design of business capabilities by using enterprise modeling techniques as a starting point, and to employ model-based patterns to describe how the software application can adhere to changes in the execution context. -------------------- • The proposal has a strong business orientation. • It was our intention to give a contribution to a gap between
business requirements and mainly technical-driven solutions for the cloud.
• With the proposed approach business modelers will be able to plan the business requirements for the cloud using goals and KPIs in accordance to changing contexts.
• The solution has been driven by empirical requirements that we have experienced so far, in particular the presented EnRiMa project case.
Future Work: FP7 CaaS project
Future Work: FP7 CaaS project
Contact Info
Jelena Zdravkovic, [email protected]
Janis Stirna, [email protected]
Martin Henkel, [email protected] Department of Computer & Systems Sciences (DSV), Sorckholm University, Seden
Jānis Grabis, [email protected]
Information Technology Institute, Riga, Technical University, Latvia